Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Municipal Finance Software topic
No spam. Unsubscribe anytime.
Council hears demo of Scenario financial‑modeling tool; staff ask about AI accuracy and security
Summary
Norristown officials watched a demonstration of Scenario, a PFM Financial forecasting platform, explored a proposed $20,000 annual subscription and asked about AI accuracy, data security (hosting in AWS) and implementation timelines; presenters said a live model can be ready rapidly and emphasized verification of AI summaries.
Norristown Municipal Council spent part of its June 16 work session reviewing Scenario, a cloud‑based financial‑forecasting and modeling platform presented by PFM Financial representatives Jim Arms and Melissa Mays.
The presenters described Scenario as a tool to run multi‑year "what if" scenarios, produce dashboards and show impacts on borrowing power (Moody's scorecard) and operating results. They demonstrated an audit trail feature that lets users drill from a summary number back to source entries and described an AI‑driven "explain" function that summarizes selected figures.
Council members probed AI accuracy and data security. One council member cited that the AI partner (Claude) can have an "about 4 percent hallucination rate" when extracting values from complex documents; presenters responded that AI is used for summaries while engineered calculations and logic are set and verified by systems engineers. Presenters emphasized that the tool is an assistant and that users must verify outputs. On security, PFM staff said Scenario is hosted in Amazon Web Services and supports role‑based access and two‑factor authentication; they can tailor access per department (for example, limiting the police chief to only a police budget view).
On procurement and operations, presenters proposed a 90‑day implementation timeline with a live model typically available within the first week after data import, weekly one‑hour meetings during onboarding, an export function for client data (CSV/Excel/PDF), and a goal of completing implementation within 90 days. Pricing cited during the presentation was about $20,000 per year for the configuration shown, with first‑year setup included; the presenter offered to lock a quoted price for three years. The council discussed whether the tool would save staff time and whether it integrates with ERPs via APIs; presenters said they can import data by spec and build an API if needed but often start with exported spreadsheets.
Why it matters: Council members said long‑range fiscal scenario planning will aid budget decisions on major capital projects discussed later in the session. Presenters and staff agreed that while AI summaries are a time‑saver, council and staff must maintain verification procedures.
What’s next: Staff indicated interest in follow‑up demonstrations and a potential Zoom meeting for more detailed review and pricing confirmation.
